How to Create a Human Voice UI

Voice UI is a hot topic, and one that brings with it a lot of assumptions. Much as Star Trek imagined the first iPad back in the late 1980s, films and television today suggest Voice UI could become a friend, personal assistant, and more. But in reality we’ve a ways to go.

For starters, voice UI needs your voice and tone. But you can’t just use your typical brand voice, because what comes across well while written doesn’t feel the same when spoken. Try reading this paragraph aloud and you’ll see what I mean. What sounds casual reads as careless, and what sounds formal reads as smooth.

What Makes Voice UI Successful?

Jean Rosenthal, the famed Broadway lighting designer responsible for the original West Side Story production, once said that good lighting shouldn’t call attention to itself.

…when no one in the audience knows where the light on the stage comes from, and when no one notices anything on the stage except the actors, the sets, the costumes[…] then you know that you have done your job as it should be done.

Jean Rosenthal, Broadway lighting designer

Thomas Gayno, Product Lead at Spotify, said something similar about voice UI:

Tomorrow’s devices should be unobtrusive… something so “you” that it dissolves into your life.

Thomas Gayno, Product Lead at Spotify

In other words, voice UI is a tool. And much like any other tool, it needs to be intuitive, and help people accomplish their tasks.

How Does Voice UI Differ from Written Text?

There are a few key ways in which voice UI is different from written text. These include:

  • Speak it aloud: natural language sounds different. Read what you’re writing and imagine saying it to another person. Now say it without looking at it – did you change your contractions?
  • Keep it short: in written text, someone can look back or reread. With voice, your user is relying on memory.
  • Don’t assume context: on a webpage or app there’s navigation, breadcrumbs, headers, and other context to tell someone where they are and how they came to be here. Keep in mind that none of this context exists with voice!
  • Accessibility takes on new meaning: with a written experience, accessibility mostly relies on metadata. For voice, accessibility require paying attention to accents, cultural phrasing, understanding lisp, and offering options before users get frustrated or overwhelmed.
  • Lastly, voice UI sounds different: NPR even found that their voice-driven application required a different set of tones, because there were different scenarios than in the radio shows.

Focus on Communication

As with any exciting new technology, it’s easy to focus on what we might someday do. But voice UI is a tool. We need to use it to accomplish goals – and we need to ensure we’re communicating in a human way.

Did you find this article useful? Share it!

Leave a Reply

Your email address will not be published. Required fields are marked *